Process mining is a recent set of techniques for analysing business processes. It is most famous for discovering how real-life processes look like based on event logs, but it also allows for checking the conformance between the de facto models discovered and any de jure models on hand, as well as for investigating process performance. Unfortunately, the contemporary process mining techniques make many assumptions about data structure and quality, at the expense of practical aspects such as usability. This thesis contributes to filling that gap by creating a framework for trying out and adopting process mining techniques in organisations. The framework consists of four steps – (1) justifying the process mining effort, (2) transforming the available data into process-aware form, (3) deriving process features as output of the process mining effort, and (4) deploying and analysing the derived process features. It is produced and demonstrated in a case study where the logs of Computed Tomography (CT) scanners are analysed in an attempt to innovate the customer service of the CT device manufacturer Siemens. The results suggest that the framework is a good way to adopt process mining techniques.